EUVL (extreme ultraviolet lithographys) is expected to succeed Deep UV (ultraviolet) lithography technology for the production of silicon-based computer chips, at and beyond the 35 nm node. This technology is expected to take over fabrication in the 2007-2009 time frame. The stepper machines that print these chips are expected to cost $20-40 M each, and, in this time frame, anticipated sales of 200-300 units/year are expected, providing the three major stepper manufacturing companies, ASML (Netherlands & USA), Nikon and Canon (Japan), with a new $100 B/year market. The light sources for these steppers, are currently required to provide greater than 100 W of ‘clean power’ and can account for up to 20% of this total market. A source of sufficient power is identified as the principal problem area in the ITRS (SEMATECH) Roadmap for the development of EUVL. The roadmap has been modified periodically over the years to take into account the required increase in wafer throughput, larger (300 mm) wafers, and higher Cost of Ownership (CoO), and the power of the source demanded has progressively increased. Currently the total required emitted power within a solid angle of 2π, from a source of <1 mm in size within a 2% bandwidth at a wavelength of 13.5 nm, is 400-1000 W. This large amount of power is the major challenge for companies developing the light sources.
There are two primary types of light sources being developed, those that depend on electrical discharge plasma, and those that use a laser-plasma source. Both approaches operate at frequencies in excess of 6 kHz, with pulse-to-pulse stability of approximately 1%. They are also required to be capable of long term operation (up time >95%), and ‘clean’ operation. By ‘clean’ operation we mean ‘debris-free’ or protected from the effects of particulate emission and plasma ions emanating from the source.
Both laser plasmas and discharge plasmas can produce high velocity particulate emission or ‘projectiles’ that will damage the expensive, precision-coated EUV collection mirrors that are in direct line-of-sight of the source. In laser plasmas, this particulate debris can originate from solid target sources, or close-proximity nozzles used to inject gaseous targets. In discharge sources the debris originates from the electrodes or from insulative materials close by. The plasma ions are, of course, inherent to the plasmas themselves. They need to be stopped from sputtering (albating) the collection mirrors. Several techniques have been devised to stop the sputtering, including Repeller Field approach disclosed in U.S. Pat. No. 6,377,651 issued to Richardson, et al. on Apr. 23, 2002, which is incorporated by reference.
Companies developing discharge plasmas (DP) are Phillips (hollow-cathode discharge), Xtreme Technologies (HC Z-pinch), Cymer (dense plasma focus), PIEX LLC (star discharge), EUVA (capillary discharge pinch plasma). Those developing laser plasma (LP) sources are Grumman Technologies (TRW and CEO), Xtreme Technologies, JMAR Research Inc, EUVA (Ushio and Komatzu), Exulite (France), Powerlase (UK) and Innolite (Sweden).
Most of these companies are focusing their R&D activities on Xenon-based plasmas. Those using discharges are injecting gaseous Xenon into the source. Those developing LP sources use ether a high-pressure Xe gas spray, which in a vacuum generates Xe nano-particles or ‘clusters’, or they are using thin (approximately 20 μm diameter) liquid Xe jets as targets.
Although the use of Xenon mitigates, to some extent the debris problem, its principal drawback is low conversion efficiency to in-band, 13.5 nm EUV light. Both DP and LP sources have been limited to conversion efficiencies (CE) of 0.5-0.7%. The highest CE recorded so far has been 0.95%. Moreover, there are now solid, atomic physics, reasons to believe that the CE of Xenon will not improve much beyond these values. These low CE's have adverse implications for both DP and LP sources. For LP sources it means the use of laser systems having a power in excess 40 kW, beyond current technical capabilities and possibly prohibitively expensive. For DP sources, the low CE poses extreme problems with heat removal from the source and very large electrical power requirements (approaching 1 MW).
The present invention advances the art by inclusion of a method, apparatus and system for generating a known number of nanoparticles of various substances, from a material-specific source, irradiated by a low power pulse laser source. There are a wide number of possible applications for use of the novel method, apparatus and system, ranging from pulse laser deposition (PLD) techniques to the generation of small numbers of nanoparticles across a specific area for biological or biochemical applications. We cite here one specific application of this generator, as a component in a system for high-power short wavelength incoherent light sources for applications in EUV lithography, advanced microscopy and precision metrology.